【目的/意義】將人工智能技術與可視化技術相結合,解析文章題名與摘要中包含的隱藏信息,為探索我國圖書情報領域的研究歷史與現狀提供全新研究視角。【方法/過程】從CNKI數據庫中下載CSSCI來源期刊中圖書情報學領域的全部文章元數據148 956條,利用自然語言處理技術抽取題名與摘要中的標簽、分類以及情感傾向,通過統計、聚類、神經網絡預測以及知識圖譜等方法,繪制期刊下載與被引關系圖、構建影響文獻被引因素模型、研究熱點與情感傾向遷移圖以及高質量文章的知識圖譜等。【結果/結論】揭示了1957-2018年圖書情報領域文獻被引與下載的關系,挖掘文獻被引的影響因素,找出研究熱點的變遷演化以及高被引文獻的特征。

【Purpose/significance】Integrating artificial intelligence technology with visualization technology, this paper analyzes the hidden information contained in the title and abstract of the article, and provides a new research perspective for exploring the research history and current situation in the field of Library and information in China.【Method/process】Download 148956 metadata of all articles in the field of Library and information science from CNKI database. Use natural language processing technology to extract labels, classifications and emotional tendencies in title and abstract, and drawdownloaded & cited graph of periodicals by means of statistics, clustering, neural network prediction and knowledge map, construct citation factor model, research hotspots and emotional propensity transfer map, and high-quality article knowledge map, etc. [Result/conclusion] This paper reveals the relationship between citation and downloading of documents in the field of Library and information from 1957 to 2018, excavates the influencing factors of citation, and finds out the evolution of research hotspots and the characteristics of highly cited documents.